HETDEX and PFS Eiichiro Komatsu (Max Planck Institute for - - PowerPoint PPT Presentation

Mapping the large-scale structure

• f the Universe with emission-line

galaxies from z=0.6 to 3.5:

HETDEX and PFS

Eiichiro Komatsu (Max Planck Institute for Astrophysics) OATs Seminar, Osservatorio Astronomico di Trieste July 15, 2019

Why Large-scale Structure?

• “End-to-end Test of the Universe”
• Cosmology as an initial-value problem
• The initial fluctuation is constrained quite well by the

cosmic microwave background data

• We then evolve the initial fluctuation forward, assuming

a cosmological model and gravitational theory

• Does the prediction agree with what we see in the data in a

late-time Universe?

2

State-of-the-art

• There is an indication that the E2E test is failing for a

flat ΛCDM model

• H0
• Amplitude of matter fluctuations in a low-z universe
• There is also an indication that the current large-scale

structure data sets may not be consistent with each other

3

appeared on July 12

5

How do we explain this?

• Insomma, non so come…
• This is the “Early Universe Probe vs Late Universe Probe” tension
• My approach is to “ask the sky”. We keep cross-checking them

with more data, until we find new explanation(s)

• In fact, it may not be just H0…
• The amplitude of matter density fluctuations in the late time

Universe measured by the large-scale structure seems low compared to what we infer from CMB

• Not yet too significant (~3σ), but it is persistent
• More data on both early and late Universe probes are necessary

6

2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 LiteBIRD [JFY 2027–] CCAT-prime [2021–]

CMB: Early Universe Probe

I talked about these 4 weeks ago HETDEX [2017–2023] PFS [2022–]

LSS: Late Universe Probe

Today’s topic

7

8

• The present-day amplitude of the matter fluctuation constrained

by the low-z data appears to be smaller than the one predicted by the evolution model given CMB

Amplitude of Fluctuations

Dark Energy Survey Collaboration

Amplitude of Fluctuations

• The present-day amplitude of the matter fluctuation constrained

by the low-z data appears to be smaller than the one predicted by the evolution model given CMB

HSC Collaboration

Amplitude of Fluctuations

9

Amplitude of Fluctuations

Amplitude of Fluctuations

• The present-day amplitude of the matter fluctuation constrained

by the low-z data appears to be smaller than the one predicted by the evolution model given CMB

• R. A. Burenin, arXiv:1806.03261

10

Two known low-z effects

• So, there is some evidence that the end-to-end test is
• failing. Namely:
• The locally measured H0 appears to be larger than that

predicted by the CMB+

• The locally measured amplitude of fluctuations appears to

be lower than that predicted by the CMB+

• Two effects that are known to influence

the low-z evolution:

• Neutrino mass
• Dark energy/modified gravity }

Large-scale structure!

11

Hobby-Eberly Telescope Dark Energy Experiment

Location McDonald Observatory (West Texas) Primary Mirror Size 10 m Location Subaru Telescope (Hawaii) Primary Mirror Size 8.2 m Wavelength Coverage 350–550 nm (Δλ=6.2Å) Wavelength Coverage Blue: 380–650 nm (Δλ=2.1Å) Red(LR): 630–970 nm (Δλ=2.7Å) Red(HR): 710–885 nm (Δλ=1.6Å) NIR: 940–1260 nm (Δλ=2.4Å) Redshift (Lyα) z=1.9–3.5 Redshift ([OII]) z=0.02–0.74 z=0.69–1.60 z=0.90–1.37 z=1.52–2.38

PFS

Spectrograph Type Integral Field Unit (IFU) # of fibers 34,944 Spectrograph Type Robotic Multi Object Fiber-fed # of fibers 2,394 + 96 Field of View 0.1 deg2 (22’ diam.) Field of View 1.25 deg2 (1.38 deg diam.) Fiber Diameter 1.5 arcsec Fiber Diameter 1.2 arcsec Survey Type Blind Survey Type Traditional Survey Volume 8.2 (Gpc/h)3 Survey Volume 2.8 (Gpc/h)3

~20 Mpc in one go!

Hobby-Eberly Telescope Dark Energy Experiment PFS

Texas-led $42M experiment Japan-led $85M instrument

Three major science programs:

• Cosmology
• Galaxy Evolution
• Galactic Archeology

But, we can do:

• Properties of Lyman-alpha emitting galaxies
• Blind survey: Unbiased survey of everything

Main Objective: Spectroscopic follow-up of targets detected by the imaging survey of Hyper Suprime Cam Main Objective: Cosmology CPPC

NEPG

13

Large Redshift Lever Arm: One Example

• We want accurate and robust cosmology! (not just precision)

PFS Collaboration Addison et al. (2018)

14

Science Cases (Cosmology)

• Not just testing tensions in H0 and the amplitude of

fluctuations!

• To rule out the standard ΛCDM model (or to put the

tightest limits on deviations)

• If ΛCDM, HETDEX can detect Λ at z>2 for the first time
• To rule out the inverted hierarchy of the neutrino mass

(or to discover it)

• And, we do a lot of non-cosmological projects too!
• I would love to discuss other ideas with you today.

These instruments are really amazing

15

Experimental Landscape

2019 2020 2021 2022 2023 2024 Euclid (launch sometime in Jan-June 2022?) DESI: 500 nights

PFS: 300 nights

2025 2026 2027 commissioning comm.

launch window

HETDEX

16

Experimental Landscape

• We are the only players at z>2.

Lasting impacts well beyond Euclid (~a billion dollar mission)

PFS Collaboration

17

Experimental Landscape

• We are the only

players at z>2

• Lasting impacts

well beyond Euclid (~a billion dollar mission)

Ariel Sánchez

18

Long fibers! (Each fiber sees 1.5”) Put into cables... One IFU feeds two spectrographs 448 fibers per IFU A test IFU being lit

IFUs fabricated at AIP

19

Current VIRUS

• 47 IFUs (out of 78) are active now. More IFUs will be installed

as they are built (at the rate of 3 units per month)

• 47 x 448 = 21,056 fibers! And this is the open-use instrument

HETDEX Collaboration

*VIRUS = Visible Integral-field Replicable Unit Spectrograph

Example of full field on M3. Green boxes are the IFU locations.

Karl Gebhardt

21

~1 arcmin, completely filled by fibers (after 3 dither)

Prime Focus Instrument (2 tons!) Fibers Detectors / Cryogenic system

Hobby-Eberly Telescope with VIRUS

One VIRUS Detector Unit cameras

Tracker (“An eye ball”)

23

This is the real one!

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 90 80 70 60 50 40 30 20 10 −10 −20 −30 −40 −50 −60 −70 −80 −90

COSMOS GOODS−N GOODS−S EGS UDS SDSS DR7

HETDEX main extension

HETDEX Foot-print (in RA-DEC coordinates)

One exposure is 20 minutes

300 deg2 150 deg2 Volume = 2.8 (Gpc/h)3 450 deg2

25

Survey Strategy

4000 shots in the northern region (“spring field”)

26

• Each “shot” in the sky contains 78 IFUs
• Spending 20 minutes per shot ~ 200 LAEs
• We do not completely fill the focal plane
• This is the “sparse sampling” technique

Sparse sampling paper: Chiang et al. (2013)

27

• λ=350–550nm with the resolving power of R~700

down to a flux sensitivity of a few x 10–17 erg/cm2/s will give us:

• ~1M Lyman-alpha emitting galaxies at 1.9<z<3.5
• 1/10 of them would be AGNs
• ~1M [OII] emitting galaxies at z<0.47
• ...and lots of other stuff, like white dwarfs,

blindly selected/discovered

What do we detect?

28

Current HETDEX Data

• 64 million calibrated spectra!
• 47,880 IFUs on the sky
• 47,880 x 448 (# of fibers per IFU) x 3 (dither) = 64M
• And this is only 10% of the full survey data!
• Goal: 468,000 IFUs on the sky
• 629M calibrated spectra. This is the big data!

(~10% of the full survey data)

29

14

A typical hetdex field

Reconstructed image of the 21k fibers. Filled squares are active IFUs,

• pen squares are those

remaining. In this frame, we would use about 15 of the stars for astrometry and throughput measures.

Karl Gebhardt

Example calibration check, using 2 white dwarfs from SDSS (virus in red, SDSS in black)

Karl Gebhardt

24

Examples from one field

Karl Gebhardt

33

SDSS-III/BOSS (z=0.6) HETDEX (z=2.5)

Full survey expectation for

35

36

One of the “Red” Spectrograph Modules being tested at LAM

One of the “Red” Spectrograph Modules being tested at LAM

by K. Yabe

39

Robotic Fiber Positioner “Cobra”

40

HSC Image of M31 (HSC FoV=1.8 sq. degrees)

reduced by HSC pipeline (Princeton, Kavli IPMU, NAOJ)

Masahiro Takada

PFS will populate 2394 individual fibers for simultaneous spectroscopy

• ver this hexagonal field.

~1.5 deg

Masahiro Takada

42

PFS Foot-print (in RA-DEC coordinates)

1400 deg2 Volume = 8.2 (Gpc/h)3

43

PFS Foot-print (in RA-DEC coordinates)

• verlap

Great region for cross-checks: LAE and [OII] in z=1.9-2.4

44

Target Selection

Shun Saito

Number of emission-line galaxies predicted by “COSMOS Mock Catalog (CMC)”

Goal: To select objects in 0.6<z<2.4 from the galaxies detected by HSC

45

Target Selection

Shun Saito

Number of emission-line galaxies predicted by “COSMOS Mock Catalog (CMC)”

Goal: To select objects in 0.6<z<2.4 from the galaxies detected by HSC

46

Example Deliverables: Galaxy Power Spectra

• There are six more redshift bins

Ryu Makiya

47

PFSxHSC: Galaxy-weak lensing Cross Spectra

S/N=19 S/N=26 S/N=23 S/N=19 S/N=12 S/N=8 S/N=6

Ryu Makiya

PFSxHSC: Testing gravity

Pengjie Zhang

PFS’s unique territory

Galaxy-shear correlation divided by galaxy clustering

General Relativity “DGP” braneworld

49

Neutrino Mass

50

Two major goals

• To rule out the inverted mass hierarchy of neutrino masses

by measuring ∑mν < 0.1 eV (95% CL)

• or, to determine the total mass if ∑mν > 0.1 eV
• To rule out the ΛCDM model by finding time evolution of

dark energy density, ρDE = ρDE(t) ≠ Λ

• or, to confirm it with unprecedented precision to z=3.5

51

Why Neutrino Mass Hierarchy?

• We know that neutrinos have masses, but we do not

know the absolute value of the mass

• Only mass differences between three mass eigenstates

are known from the neutrino oscillation experiments

• Knowing the mass would be nice, but what appears to be

more fundamental is the mass hierarchy

• “Normal” vs “Inverted”

52

Mass Hierarchy

• Do we have two heavy states (inverted), or just one

(normal)? ∑mν = 0.1 eV is the key level

From Patterson (1506.07917)

53

Are neutrinos Dirac or Majorana?

• Deciding the mass

hierarchy sets a concrete target for the neutrino-less double beta decay experiments

• Dirac or Majorana?

Fundamental importance!

Capozzi et al. (2016) inverted normal

54

Neutrino Mass Target in Landscape

• The current upper bound from cosmology

(Planck+BOSS): ∑mν < 0.16 eV (95% CL; Alam et al. 2017)

• Planned laboratory (i.e., non-cosmological) neutrino

experiments would yield:

From Patterson (1506.07917)

Effects of Massive Neutrinos in Cosmology

• Neutrinos do two things:
• 1. Change the expansion history of the Universe
• 2. Slow down the structure formation

BAO, AP RSD, Shape

by C. Hikage, R. Makiya, A. Sanchez, N. Sugiyama [68%CL] [95%CL]

If the total neutrino mass is ∑mν=0.06 eV

But, what about cosmological model-dependence?

58

Neutrino mass from cosmology: Model dependence!

• A typical thing you see at conferences:
• A cosmologist: “So, this is our measurement of the

total neutrino mass from cosmology. This is much better than the laboratory experiments!”

• A particle physicist: “Nice, but how dependent is your

constraint on the assumed cosmological models”

• A cosmologist: “Ah… Um…”
• Let’s settle this!

59

Deconstructing the neutrino mass constraint from galaxy surveys

• Neutrino mass changes:
• Expansion rate (hence distances)
• Overall growth of matter fluctuations
• But, these effects can be

mimicked by other effects in cosmology

• The scale-dependent suppression of

the power spectrum is not!

• This is unique to neutrino masses

(in General Relativity)

Boyle & Komatsu (2018); Boyle (2019) Aoife Boyle (MPA)

60

Distance Effect

[Forecast for Euclid]

Strong dependence on the assumed cosmological models!

Boyle & Komatsu (2018); Boyle (2019)

61

Overall Growth Effect

[Forecast for Euclid]

Strong dependence on the assumed cosmological models!

Boyle & Komatsu (2018); Boyle (2019)

62

Free-streaming Shape Only

[Forecast for Euclid]

Model dependence disappears!

Boyle & Komatsu (2018); Boyle (2019)

63

Everything Combined

[Forecast for Euclid]

Constraints tighten, but the model dependence re-appears

Boyle & Komatsu (2018); Boyle (2019)

64

Precision vs Robustness

• If we want precision, we may combine all the information

and report the neutrino mass constraint

• But, we must be honest and admit the cosmological

model dependence!

• If we want robustness, we can get a model-independent

constraint on the neutrino mass from the free-streaming signature in the power spectrum

• Particle physicists would be happ(ier)?

65

Summary

• Galaxy surveys are going to high redshifts
• PFS: 0.6<z<2.4; HETDEX: 1.9<z<3.5
• Blind nature of HETDEX is very exciting for new discoveries
• Checking for the internal consistency of ΛCDM over a wide

redshift range. Is the H0 tension due to low-z effect?

• Measurement of the neutrino mass may be “just around the

corner”

• But beware the cosmological model dependence. The free-

streaming signature is a promising way to remove the model dependence

66

Final Message

• Both instruments are open-use!
• VIRUS (IFU used by HETDEX) is publicly available
• PFS will be publicly available also after ~2022
• Use them! I would be very happy to talk with you about

new ideas

67